[Therion] Duplicates including adding splays?
This bug should be also fixed now in 5.3.8. If not, please let me know.
S.
On 13 June 2011 10:00, Stacho Mudrak wrote:
> It looks like a bug, they should not be used for depth. I will try to fix
> it.
>
> Thanks, S.
>
> On 13 June 2011 08:28, marco corvi wrote:
>>
>> I agree with bruce.
>>
>> also noticed that splay shots are used for the "depth".
>> it might make sense, but  i rather have the depth
>> computed with the survey shots only.
>>
>> marco.
>>
>> On Mon, 2011-06-13 at 08:44 +1200, Bruce wrote:
>> > Hi Andrew
>> > Yes, when splay shots to an unlabelled station such as '-' are included
>> > in a
>> > dataset, such as when using a distoX or similar, the statistics do not
>> > make
>> > that much sense to me. I think it still works the way you describe in
>> > 5.3.7.
>> >
>> > To my way of thinking, splay shots are not part of the survey length
>> > (and
>> > therefore should not be duplicate either), they are analogous to LRUD.
>> > Â IF
>> > they are to be reported, they should be separately reported as 'total
>> > splay
>> > length' and 'number of splay stations' (could be further broken down
>> > into
>> > 'surface splays' and 'cave splays' if one was being picky).
>> >
>> > If there were to be a compromise for the sake of simplicity I would
>> > prefer
>> > that therion exclude splay information from all the reported statistics
>> > (as
>> > it ignores LRUD, which are really just special splay shots), rather than
>> > the
>> > apparently selective way the splay shots are currently included.
>> >
>> > Regards
>> > Bruce
>> >
>> > -Original Message-
>> > From: therion-bounces at speleo.sk [mailto:therion-bounces at speleo.sk] On
>> > Behalf
>> > Of Andrew Atkinson
>> > Sent: Monday, 13 June 2011 6:10 a.m.
>> > To: List for Therion users
>> > Subject: [Therion] Duplicates including adding splays?
>> >
>> > Hello
>> >
>> > Therion 5.3.5 ubuntu 10.04
>> >
>> > In the attached file the following information is given about the survey
>> >
>> > title: RhinoPassage
>> > length: 189.07m (surface 0.00m, duplicated 24.58m)
>> > vertical range: 20.51m (from 1.27 at RhinoPassage at 20.38m to
>> > 1.3 at RhinoPassage at -0.13m)
>> > north-south range: 92.47m (from 1.27 at RhinoPassage at 33.43m to
>> > 1.53 at RhinoPassage at -59.04m)
>> > east-west range: 66.67m (from 1.53 at RhinoPassage at 66.67m to
>> > 1.0 at RhinoPassage at 0.00m)
>> > number of shots: 281
>> > number of stations: 282
>> >
>> > The duplicate figure got me wondering, it seems to be counting the
>> > splays but the length does not. As the duplicate none splays are only
>> > about 10m.
>> > Maybe this is what is wanted, but I personally think that the duplicate
>> > value should be the amount taken off the length by the duplicate flag?
>> > Ie why is length treated differently to duplicate (something to do with
>> > the way survex treated splays and duplicate as different, I think that
>> > splays are not included in extend in survex)
>> >
>> > Maybe 2 figures would be better, duplicate length and duplicate splays.
>> >
>> > {probably best to ignore this next bit}
>> > Oh this leads to all sorts of thinks that are really not important, like
>> > maybe giving the length of all the splays, and the number of stations,
>> > is it really 282 or really 54. All depends on your definition I guess
>> >
>> > Andrew
>> >
>> > ___
>> > Therion mailing list
>> > Therion at speleo.sk
>> > http://mailman.speleo.sk/mailman/listinfo/therion
>>
>>
>> ___
>> Therion mailing list
>> Therion at speleo.sk
>> http://mailman.speleo.sk/mailman/listinfo/therion
>
>
[Therion] xtherion -close on
But in therion, syntax is different. It is: POINT CP1 CP2 POINT And I am not sure, whether therion accepts CP specification before the first point. S. On 20 June 2011 10:15, marco corvi wrote: > On Sat, 2011-06-18 at 11:07 +0200, Stacho Mudrak wrote: >> This bug should be fixed in 5.3.8. But there is one issue. If you have close >> bezier spline, how do you specify control points, if last point is not same >> as first one? >> > not sure ... > however the first point has a fore control point and a back > control point, the last point too. > i 'd expect that the piece of line between the last and the first > point uses the fore CP of the last and the back CP of the first. > > marco > > > > > ___ > Therion mailing list > Therion at speleo.sk > http://mailman.speleo.sk/mailman/listinfo/therion >
[Therion] Bug in 5.3.7
OK, next time I will create a separate message, when new snapshot will be released. S. On 20 June 2011 17:34, Wookey wrote: > +++ Stacho Mudrak [2011-06-20 17:07 +0200]: >> I have tried to implement all your fixes from debian repository. Is there any >> other way to help you with debian packaging? > > Oh, and I didn't see a '5.3.8 has been uploaded' message. It's only > because of Graham's post that I noticed. I guess you would have sent > one after he'd reported back. > > But the point is that I need notification of new releases (not that I > will necessarily upload quickly, but I definately won't if I don't > know :-) > > Wookey > -- > Principal hats: Â Linaro, Emdebian, Wookware, Balloonboard, ARM > http://wookware.org/ > ___ > Therion mailing list > Therion at speleo.sk > http://mailman.speleo.sk/mailman/listinfo/therion >
[Therion] therion 5.3.8 (was re: Bug in 5.3.7 )
> One thing I did notice: > therion.ini in 5.3.8 has unix lineends > therion.ini in 5.3.7 has dos lineends > > was that a deliberate change? It broke the > 80remove-czech-fonts-default.patch and more importantly it causes > everyone upgrading to get a 'you changed therion.ini - do you want to > use new or old' and if you ask for a diff it shows every line (which > have technicaly changed, but only in lineend). No it was not. I did not know, it will cause such problems. I am sorry. I can put it back to dos lineend in the next snapshot. I usually do changes on windows, so from time to time, it happens and .ini files were not listed to be converted to unix format. But I think, they should have unix lineends. S.
[Therion] therion 5.3.8 (was re: Bug in 5.3.7 )
+++ Stacho Mudrak [2011-06-21 18:16 +0200]: > > One thing I did notice: > > therion.ini in 5.3.8 has unix lineends > > therion.ini in 5.3.7 has dos lineends > > > > was that a deliberate change? It broke the > > 80remove-czech-fonts-default.patch and more importantly it causes > > everyone upgrading to get a 'you changed therion.ini - do you want to > > use new or old' and if you ask for a diff it shows every line (which > > have technicaly changed, but only in lineend). > > No it was not. I did not know, it will cause such problems. I am sorry. > I can put it back to dos lineend in the next snapshot. I usually do > changes on windows, so from time to time, it happens and .ini files > were not listed to be converted to unix format. But I think, they > should have unix lineends. Yes. I think it should stay linux, and I've already fixed-up the patch. And actually, so long as the file hasn't changed, the auto-update system shouldn't ask about it. I need to work out why it did for me last night. So actually this should be fine, and I'll upload soon. Don't change it back. Wookey -- Principal hats: Linaro, Emdebian, Wookware, Balloonboard, ARM http://wookware.org/
[Therion] Lox model walls all inside out when a scrap contains a loop
Thanks Bruce, With that option and a lot of messing about I finally managed to get my entrance maze to appear correctly on model! It seems that the model building for KML and LOX is very sensitive to loops. Too many and it just makes a complete mess of it. But at least now I have managed to split my maze in half and get the first half to appear correctly in both KML and LOX. Now to try and get the other half in a second scrap to work and to join them back together in a single model. Footleg On 20 June 2011 21:24, Bruce wrote: > ** ** > > Yes, > > The âinternalâ wall needs to have the option -outline in when it is all > part of the same scrap. > > Bruce > > ** ** > -- > > *From:* therion-bounces at speleo.sk [mailto:therion-bounces at speleo.sk] *On > Behalf Of *Footleg > *Sent:* Tuesday, 21 June 2011 12:56 a.m. > *To:* **List for Therion users** > *Subject:* [Therion] Lox model walls all inside out when a scrap contains > a loop > > ** ** > > I cannot get a lox model to show the walls properly when I add a loop in my > scrap. I've attached the files I have been messing about with to test this. > There is one loop in the scrap sketch (created by line 9, the last line in > the file). As provided this does not display properly in **Loch**. But if > I delete that last line so that the the centre of the loop is removed then > the loch model looks fine. Am I missing a trick here? I cannot split every > loop in my caving into separate scraps (before someone suggests that). I > have too many of them in my real survey data. > > Testing with the 5.3.8 build. > > Footleg > > ___ > Therion mailing list > Therion at speleo.sk > http://mailman.speleo.sk/mailman/listinfo/therion > > -- next part -- An HTML attachment was scrubbed... URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20110621/759396ee/attachment.html>
[Therion] No output in loch 5.3.3 Ubuntu 11.04 (Or aven for that matter)
On 21/06/11 02:54, Wookey wrote: > +++ Andrew Atkinson [2011-06-20 18:16 +0100]: >> After a hard drive failure I upgraded to ubuntu 11.04 and the default >> versions came in. therion 5.3.5 loch 5.3.3 (survex 1.1.14) > > Are you sure you have different versions of therion and therian-viewer > (loch)? This page says they should be the same for natty: > http://packages.ubuntu.com/search?keywords=therion&searchon=names&suite=natty§ion=all > > (5.3.3-1ubuntu1 for both). I did install manually from one of your web site, cannot remember which one, but pretty sure that only was xtherion the viewer did not work. I had checked in Synaptic, to see if I could force a version (A skill I have not tried yet) but it only listed one version. Screen shot attached > > I'm not sure if mismatched versions is a problem here - I guess things > should work anyway > >> I think this is the syslog line that happens when trying to open a lox, >> is there another log file I should be looking for. Nothing appears when >> opening a .3d >> >> Jun 20 17:55:38 janda kernel: [ 941.228479] loch[2762]: segfault at >> 28d02ed ip 7f53c50f6b28 sp 7fff21d93e90 error 6 in >> libwx_baseu-2.6.so.0.3.1[7f53c509b000+10c000] > > That seems plausible. something horrid going wrong in the bowels of > wxwindows libs. That may just be loch tickling a bug, rather than > there actually being anything wrong with loch itself. > > I failed to find any matching bug already reported in Ubuntu or Debian. > >> This is the system information that you can get out of the help menu on >> Aven (maybe it is relevant to loch) >> >> Linux 2.6.38-8-generic x86_64 >> wxWidgets 2.8.11 (GTK+>= 2.6) >> Display Depth: 24 bpp (colour) >> OpenGL 2.1.2 NVIDIA 270.41.06 > > I guess that means that you are using the proprietary nvidia drivers? > > You could see if the free drivers work any better for you (noveau). > (They tend to crash my work natty X-server on a regular basis unless > one turns turn acceleration off). Have tried it before and it was very crash prone, also did not seem to do the right settings for my monitor, not that the nvidia drivers are much better, took me ages to get it to do the right resolution, so do not really want to play that game again. > > Finding out if therion 5.3.7-1 exhibits the same problem or not is of > course interesting. That is currently avaialble for oneiric. But what is the easiest way to do this without breaking my only machine? Oneiric is still in very breaky stage! > > I've prepared a 5.3.8-1 now too. I'll put that in my PPA soon, pending > resolution of the therion.ini linefeed issue. That seems a safer option from an end users point of view , thanks :) Andrew > > Wookey -- next part -- A non-text attachment was scrubbed... Name: Screenshot.png Type: image/png Size: 14125 bytes Desc: not available URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20110621/d0d741ce/attachment.png>
[Therion] Lox model walls all inside out when a scrap contains a loop
Yes, The 'internal' wall needs to have the option -outline in when it is all part of the same scrap. Bruce _ From: therion-bounces at speleo.sk [mailto:therion-boun...@speleo.sk] On Behalf Of Footleg Sent: Tuesday, 21 June 2011 12:56 a.m. To: List for Therion users Subject: [Therion] Lox model walls all inside out when a scrap contains a loop I cannot get a lox model to show the walls properly when I add a loop in my scrap. I've attached the files I have been messing about with to test this. There is one loop in the scrap sketch (created by line 9, the last line in the file). As provided this does not display properly in Loch. But if I delete that last line so that the the centre of the loop is removed then the loch model looks fine. Am I missing a trick here? I cannot split every loop in my caving into separate scraps (before someone suggests that). I have too many of them in my real survey data. Testing with the 5.3.8 build. Footleg -- next part -- An HTML attachment was scrubbed... URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20110621/ec703c8c/attachment.html> -- next part -- A non-text attachment was scrubbed... Name: RianoEntPlan_test.th2 Type: application/octet-stream Size: 5826 bytes Desc: not available URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20110621/ec703c8c/attachment.obj>
[Therion] No output in loch 5.3.3 Ubuntu 11.04 (Or aven for that matter)
+++ Andrew Atkinson [2011-06-20 18:16 +0100]: > After a hard drive failure I upgraded to ubuntu 11.04 and the default > versions came in. therion 5.3.5 loch 5.3.3 (survex 1.1.14) Are you sure you have different versions of therion and therian-viewer (loch)? This page says they should be the same for natty: http://packages.ubuntu.com/search?keywords=therion&searchon=names&suite=natty§ion=all (5.3.3-1ubuntu1 for both). I'm not sure if mismatched versions is a problem here - I guess things should work anyway > I think this is the syslog line that happens when trying to open a lox, > is there another log file I should be looking for. Nothing appears when > opening a .3d > > Jun 20 17:55:38 janda kernel: [ 941.228479] loch[2762]: segfault at > 28d02ed ip 7f53c50f6b28 sp 7fff21d93e90 error 6 in > libwx_baseu-2.6.so.0.3.1[7f53c509b000+10c000] That seems plausible. something horrid going wrong in the bowels of wxwindows libs. That may just be loch tickling a bug, rather than there actually being anything wrong with loch itself. I failed to find any matching bug already reported in Ubuntu or Debian. > This is the system information that you can get out of the help menu on > Aven (maybe it is relevant to loch) > > Linux 2.6.38-8-generic x86_64 > wxWidgets 2.8.11 (GTK+ >= 2.6) > Display Depth: 24 bpp (colour) > OpenGL 2.1.2 NVIDIA 270.41.06 I guess that means that you are using the proprietary nvidia drivers? You could see if the free drivers work any better for you (noveau). (They tend to crash my work natty X-server on a regular basis unless one turns turn acceleration off). Finding out if therion 5.3.7-1 exhibits the same problem or not is of course interesting. That is currently avaialble for oneiric. I've prepared a 5.3.8-1 now too. I'll put that in my PPA soon, pending resolution of the therion.ini linefeed issue. Wookey -- Principal hats: Linaro, Emdebian, Wookware, Balloonboard, ARM http://wookware.org/
[Therion] therion 5.3.8 (was re: Bug in 5.3.7 )
+++ Wookey [2011-06-20 16:31 +0100]: > +++ Stacho Mudrak [2011-06-20 17:07 +0200]: > > I have tried to implement all your fixes from debian repository. Is there > > any > > other way to help you with debian packaging? > > If you can tell me which Debian patches have been > incorporated for each new release it saves me having to work it out, > by which ones don't apply and then deciding if it's because it's been > included, or because things have changed. OK. I had a look and I see you've just incorporated the gcc4.5 fixes. Excellent. That will avoid future issues for people. I think you should also look at upstreaming: 30doc-fixes.patch (manpage update) 35loch-manpage.patch 40samples-fixes.patch (build issue) 50svg-logo.patch (this could be extended to use logo in various places) 82-nolang-segfault-fix.patch There is not much debian-specific in there. For the rest we should discuss whether we can avoid me maintaining patches by upstream changes or not. e.g I'm not sure of the czech-font-by-default thing is still a real issue or not. One thing I did notice: therion.ini in 5.3.8 has unix lineends therion.ini in 5.3.7 has dos lineends was that a deliberate change? It broke the 80remove-czech-fonts-default.patch and more importantly it causes everyone upgrading to get a 'you changed therion.ini - do you want to use new or old' and if you ask for a diff it shows every line (which have technicaly changed, but only in lineend). It probably should be a unix file, so maybe it's time to have this slighly annoying transition. There is probably a way for me to manage it silently in the packaging, but having to find that out stops this being a trivial upload I could do today. Wookey -- Principal hats: Linaro, Emdebian, Wookware, Balloonboard, ARM http://wookware.org/
